Of Measures and Mismeasures :

Poverty, Inequality and Population by D Jayaraj and S Subramanian (New Delhi: Oxford University Press), 2010; pp 298 + xvi, Rs 775.

BOOK REVIEW

-	--
-
-	-
–
Of Measures and Mismeasures Achin Chakraborty		consequence of the compulsions arising from poverty or a consequence of the insensitivity or irrationality of parents who exercise their preferential choice? Finding an answer to this question is extremely

he pervasive but unavoidable tendency towards specialisation in economics seems to have created a gulf in the literature. On the one hand is the literature on axiomatic measurement theories in the areas of poverty and inequality; on the other is the empirical development literature, primarily motivated by the need for making substantive conclusions on poverty or inequality in the speciﬁc context of a country or region. While much of the empirical work in this area can be criticised by theorists on the grounds that it ignores the fundamentals of normative measurement theory, pure theorists of measurement rarely pay attention to empirical issues in measurement and dirty their hands with data. And so the gulf persists.

In a series of 12 papers written over a period of about 12 years and put together in the present volume, Jayaraj and Subramanian (JS) seem to have set out to bridge this gulf, even though nowhere in the introduction do they describe such bridge making as an explicit goal. Instead of pontiﬁcating on what is to be done, they actually show how conceptual foundational issues can be combined with rigorous analysis of data in order to illuminate certain aspects

Poverty, Inequality and Population by D Jayaraj and S Subramanian (New Delhi: Oxford University Press), 2010; pp 298 + xvi, Rs 775.

of development which would otherwise be ignored. Some essays are primarily methodological; their main focus is on how one can meaningfully measure things like “group-discrimination”, “agedness” or “femaleness” of a population. When JS are not busy constructing another index, however, they sift through a pile of data to ﬁnd stories of inequality worth telling, as in the case of “The Distribution of Household Wealth in India”, which primarily aims to draw substantive conclusions on important development issues by making innovative use of Indian data.

Reorienting Habitual Modes of Thought

The ﬁrst essay on child labour is an excellent example of how an explicit recognition of the variety of practical decision problems associated with empirical work inﬂuences the way that theoretical and conceptual arguments are translated into practice. It starts with the following research question: Is the phenomenon of child labour a

October 1, 2011

important because it has direct bearing on the policy question of whether or not child labour should be banned outright.

Starting from a set of axioms, JS set up a simple model to derive the child labour supply function. The most crucial assumption in the model is that a household’s sole motive for the employment of child labour is to ensure the survival of the household. The immediate implication of their child labour supply function is that a blanket ban on child labour, without any other measure to support the affected family economically, would push the family to sub-subsistence level of income. The strength of this important policy conclusion, however, crucially depends on the empirical corroboration of the presumed relationship between the incidence of child labour and poverty.

One may fail to ﬁnd such a relationship if one takes a narrow deﬁnition of child labour that excludes the work performed in domestic chores. JS make a strong case for adopting a more expansive deﬁnition of child labour that would include children engaged in domestic work. They also strongly advocate assessment of deprivation in the multidimensional space of capability deprivation rather than that of income

vol xlvi no 40

Economic & Political Weekly

alone. With this reﬁned conceptualisation of both child labour and poverty, they are able to ﬁnd a positive empirical relationship between the two. In a sense, this ﬁnding reconﬁrms our understanding that the theoretical distinction does matter to the empirical results. In the process, JS successfully reconcile the seemingly opposing strands in the literature on the determinants of child labour.

In the second essay, JS present an index which measures the relative ease with which income poverty in an economy can be eradicated through a scheme of progressive redistributive taxation. The proposed scheme is rather simple. The richest person’s income is taxed to the point where her income is equalised with the income of the next richest person. If the tax amount is sufﬁcient to meet the aggregate poverty gap, nothing further is needed. But if the amount falls short of the total deﬁcit, the next richest person is taxed to the point where her income is equalised with the income of the third richest person, and so on. It is an exercise in pure normative assessment, abstracting from all the positive economic implications, such as incentive effects, of such radical income transfers.

Given the practical political economy kind problems in implementing such a radical scheme, one might be sceptical about the real worth of the exercise. However, the fact that empirical estimation of the “index of difﬁculty of eradicating poverty” (p 69) throws up a rather low value for India can potentially exert moral pressure on policymakers by pointing out in clear terms the relative ease with which poverty can be eradicated. We had a rough sense of this from the recurring presence of several Indians in the Forbes billionaires list. We now know for sure that not enough income to go around can hardly be a legitimate excuse.

Discrimination and Disadvantage

The issue of inter-group disparity in poverty and well-being is taken up in the next two essays. In the third chapter this is called “groupdiscrimination”; the fourth essay calls it horizontal inequality. Prasanta Pattanaik found “discrimination” too loaded with its connotation of “deliberate unfair treatment arising from prejudice”, and suggested “relative

BOOK REVIEW

d isadvantage” in its stead (p 77). JS appreciated the suggestion but went ahead with their chosen term. In their own words (p 77):

[i]f one assumes – as we do throughout this paper – that whenever one group is worse off than another it is undeservedly worse off, then the relative disadvantage experienced by the group can be construed as amounting to discrimination against it.

Even though most welfare economists (Pattanaik being one of the most eminent) might be a bit reluctant to accept this assumption, scholars in other

social sciences are more likely to be favourably disposed to this view. Kaviraj (1996: 115), for example, writes

Disadvantage is seen more as unjust treatment of whole communities, like lower castes, minority religious groups and tribal communities, which are thus seen as potential political actors for social equality….Certainly, people who are part of democratic mobilisations are predominantly poor, but the principle of their self-identifying action is not poverty but discrimination.

No matter what we call it, horizontal – or inter-group – inequality is no longer as neglected a dimension as it was a decade

BOOK REVIEW

ago (JS quote Frances Stewart’s 2001 claim to this effect at the beginning of Chapter 4). At least in India, researchers routinely produce comparative analyses of the relative status of scheduled castes (SCs), scheduled tribes (STs) and others, based on descriptive statistics on a variety of quantiﬁable aspects of well-being and development. The work of JS stands quite apart from the commonplace applied work of this kind. It has been instrumental in facilitating methodological thought on quantitative presentation of systematic group-related disparities in the distri bution of well-being and deprivation.

In the third essay, JS formally derive a group disparity-adjusted poverty measure and apply it to National Sample Survey (NSS) data. Not surprisingly, they ﬁnd that comparative (interstate) analysis of poverty undergoes some changes once we bring between-group disparities into the assessment of overall poverty. In India, it is well known that poverty among SCs and STs is much higher compared to others. The group disparity-adjusted poverty measure of JS combines aggregate poverty with intergroup disparity into a single measure so that it yields a unique ranking of states. This may not sound terribly interesting.

What is more interesting is the axiomatic characterisation of their measure and its implications for group justice.

The crucial axiom is “Increasing Positive Responsiveness to Subgroup Poverty” (IPR), which says roughly that a given reduction in subgroup poverty causes a greater reduction in overall poverty, the poorer the subgroup. This sounds similar to the conventional transfer axiom in the poverty measurement literature. But here it is formulated in terms of groups rather than individuals. What is interesting is that an index which satisﬁes IPR could well violate the transfer axiom. In other words, considerations of group justice could come into conﬂict with notions of individual justice implicit in standard distribution-sensitive poverty measures such as Sen’s index. JS rightly suggest that we may have to reject certain conventional axioms of poverty measurement if we are really concerned about group justice.

The fourth essay reports an interesting journey to discovery. Starting from certain very simple and intuitively plausible measures of horizontal inequality, JS go on to derive their vertical counterparts and obtain certain well-known measures of inter-personal inequality. In the process, somewhat accidentally, they hit upon a new measure of inter-personal inequality which turns out to satisfy some very useful properties.

Innovative Demography

Given information on the fertility and mortality rates of two populations, some inference on their well-being regimes could be drawn if their relative “agedness” was known unambiguously. This motivated JS to develop ﬁrst a graphical device called the binary age comparison (BAC) curve to visually compare the age structures of two populations, and subsequently to derive axiomatically a summary index of agedness. The axiomatic structure has parallels with the literature on the measurement of poverty. Continuity, focus, monotonicity, transfer – all the familiar axioms from the poverty literature enter here with appropriate interpretations to suit the problem at hand. However, not all the poverty axioms can be easily reinterpreted because of

ambiguity over underlying value judgments in the context of population ageing. For example, how should we compare ageing by one additional year for someone who is 65 years old versus someone at 75? If we keep in mind the parallel with the transfer axiom in poverty measurement, the idea that the

October 1, 2011 vol xlvi no 40 EPW Economic & Political Weekly

BOOK REVIEW

e xtent of ageing is not the same in the two cases seems reasonable. But it is not easy to settle on who, in this example, has aged more in one year – someone who has become 66 or someone who has become 76. JS are in favour of the value judgment that “a person ‘ages more’ in moving, say, from 65 to 66 years of age, than in moving, say, from 75 to 76 years” (p 160). It is not hard to imagine that others could hold the exact opposite view, i e, an older person ages more than a younger one in one year.

The possibility of diametrically opposed views of this kind is rather unsettling. How does one make connections between agedness and various dimensions of well-being then? Agedness is deﬁnitely not a constituent element of well-being. Is it a determinant of well-being? The empirical illustrations in the paper, JS caution, are only for illustrative purposes. They are all right as far as they go. But, because of the inherently ambiguous connection between agedness and general well-being, it is yet to be seen how by combining the index of agedness with different demographic and well-being indicators, substantive conclusions can be drawn with real data. Nevertheless, JS must be appreciated for forcing us to think in this direction.

Debating Sex Ratio

As a general theme, the connection between demographic indicators and population well-being runs through a couple of other essays as well. In Chapter 9, JS ﬁrst consider alternative hypotheses to explain declining sex ratios at birth (SRB) in India over a long period. Even though the female foeticide (FF) hypothesis has been doing the rounds for some time now, JS systematically argue against and debunk, with an excellent combination of fact and logic, the tendency to project FF as a “virtually monocausal explanation” (p 206). Instead, they make a strong case in support of an alternative explanation they call the “sex-neutral reduction in foetal wastage” (SNRFW) hypothesis (p 201).

The male foetal wastage rate (that includes spontaneous abortions and stillbirths) is generally higher than the corresponding female rate. If the two rates fall by the same proportion, may be due to an improvement in the well-being of the mother, the SRB should fall. Since declining SRB is also associated with a declining overall sex ratio (SR), it may therefore be misleading to see declining SR as a symptom of deterioration in women’s well-being. Having analysed a wide range of data on several dimensions of women’s well-being over time, JS conclude that in the entire period in which SRB declined, women’s well-being indeed improved. After letting logic take its own course, JS seem to realise that pure logic has not served them well. How could they be so insensitive as to claim that women’s well-being had improved? They hurriedly take off their logician shoes and let themselves slip into the comfort of radical gibe against the so-called “state-in-the-withdrawal-mode” and all the rhetoric that goes with it. JS could easily avoid this bit of irrelevancy in the latter half of the conclusion to Chapter 9.

In the next essay on “The Well-being Implications of a Change in the Sex-Ratio of a Population”, JS return to their logician selves with a vengeance. With tightly-knit logic and a modest dose of formalism, they show that it is possible in principle to have a rise in the SR as a result of the processes that diminish women’s well-being, just as it is possible to have a decline in the SR as a result of well-being improving processes. When we say decline in the SR is a bad thing, it is not clear what we mean by that. Is it bad in itself, irrespective of the well-being improving processes that could produce the decline? JS have made a very important contribution by highlighting this paradox.

The last chapter titled “Abusing Demography” is a review essay of the most serious kind. The reader can see JS at their rhetorical best. With the right combination of logic, fact, metaphor and irony, they effectively demolish the pseudo-scientiﬁc pursuit of the authors of a book they review. This powerful critique not only applies to the book under review, it also counters any such attempt to show that the numerical strength of the non-Hindu population in India is growing fast enough to push Hindus to minority status in some not-so-distant future.

A volume of this kind, consisting of papers written over a long period of time and published in different journals that differ in style and target readership, can tend unevenness, repetition and sometimes lack of relevance, being stuck with dated empirical material. The present volume stands out remarkably in this regard. The authors seem to have taken great care to avoid repetition and irrelevance and have made sure that the papers remain as readable and enriching as they were at the time of their ﬁrst appearance.

Achin Chakraborty (achinchak@rediffmail.com) is with the Institute of Development Studies Kolkata.

Reference

Kaviraj, S (1996): “India: Dilemmas of Democratic Development” in A Leftwich (ed.), Democracy and Development (Cambridge: Polity Press).